Efratom rubidium oscillator manual

John-

Thank you! I'm printing it as I type.

I purchased one of these on E-Bay several years ago, and use it to sync the horizontal sweep of an oscilloscope. I feed a 10 MHz counter timebase to the scope's vertical input, and adjust for zero drift. Frequency of the oven-controlled oscillator in an HP counter seems to hold up quite well over time.

Fred

What's amazing is to take two of these. Use one to trigger a scope, and look at the other on a vertical channel. Once things are aligned, you can crank up to, say, 10 ns/div and the rising edge looks perfectly triggered.

Come back 20 minutes later, and the edge will have moved a little. You can then calculate delta-F between the oscillators in units, like, ns per hour.

John

On a sunny day (Sat, 14 May 2011 13:57:53 -0700) it happened John Larkin wrote in :

Thats is a very interesting pdf. I notice a few things (I am no expert on atomic clock, so..) anyways, they 'wobble' at 127 Hz around Fc it seems, so for short term (< a few ms) that would give bigger errors. So one would need to drive some PLL with it I would think.

Also they use a cesium standard as calibration, is there a reason why cesium should be more accurate?

I like that manual, it reminds me of manuals I wrote many years ago (I once was technical writer for a very big [arms] company), where we had to mention every component and what it did, and how its value was calculated too. And I did that with some of my own designs, nice piece of work that manual.

Thank you for making this available.

On a sunny day (Sat, 14 May 2011 20:16:25 -0400) it happened Fred McKenzie wrote in :

xy mode should work too :-)

The rubidium resonance is used to very slowly correct the 10 Mhz OCXO. The correction is slow enough that the 127 Hz sweep doesn't show up at the output. They do a lock-in amplifier sort of loop so, when it's centered at zero error, there is no FM of the 10 MHz oscillator. I think.

I don't know the physics. The caesium standards are a lot bigger and much more expensive.

I've learned a lot from reading old manuals. Not much from newer ones.

John

Much Thanks,

George H.

Hey John, Are those your scribbles on page 80 of the manual? Trying to 'decode' the 6.834... GHz. Do you understand why only the 60 MHz signal is 'multiplied' by 114? Perhaps the 5.3 MHz is added in at a much lower amplitude so only the 60 MHz is making harmonics?

Thanks,

George H.

Yup, my scribbles.

The SRD drive is a horrible mess. And that mess is multiplied by the SRD into a forest of messes. The spectrum must be a horror, multiples of 60 MHz, each with +-5.3 MHz sidebands and worse. ONE of the zillion spectral lines hits the rubidium resonance. The rest waste power but otherwise don't matter.

There's probably an RF cavity that enhances the 6.83 GHz spectral line, its sidebands, and probably a few dozen of its neighbors.

John

Thanks for posting the scanned pages !

Jure Z.

We have a cool Sharp digital copier that will sheet feed, scan both sides, and dump the PDF onto our server. Easy!

John